I've seen a report of this study in a couple of places. Though I haven't yet seen the actual study (it doesn't seem to be published yet). Here it is from New Scientist.
Participants placed a finger in a clamp which was tightened using a pressure gauge until they reported feeling pain. Both men and women appeared to feel pain more quickly if the person turning the clamp was a man.
Previous studies have shown that men report feeling less pain in front of a female experimenters - but this was put down to their wanting to appear more macho.
David Williams, who carried out the study at the University of Westminster in the UK, says his findings contradict this assumption. Williams suggests that the subjects of his study may be socially conditioned to expect men to be more likely to inflict harm. The effect, "is likely to be a result of what participants subconsciously expect, based on socially acquired gender stereotypes," he says.
The study also shows that a person's surroundings can affect their sensitivity to pain. Objects that might be associated with suffering - such as a chart showing wounds or a poster related to blood donations - were found to make the participants report feeling pain more readily.
I'll admit I'm quite excited about this story; hopefully it will be in print soon and withstand scrutiny. Externalism about phenomenological content anyone?
[ed: This and the preceding post were old posts; I've moved them up because they are more interesting than some of the abstracts I had recently posted]